INDEX

Explanations

mentions of phrases related to legal issues, particularly highlighting instances of something being deemed unconstitutional

oai_token-act-pair · gpt-3.5-turbo

references to unconstitutional actions or practices

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pring

-0.80

ariat

-0.80

achine

-0.80

orah

-0.77

iquette

-0.77

igi

-0.74

ebus

-0.73

 Tycoon

-0.73

erm

-0.72

fleet

-0.71

POSITIVE LOGITS

 unconstitutional

0.94

 discriminatory

0.81

 unfocusedRange

0.75

 usur

0.75

 overr

0.74

 violation

0.73

 interference

0.73

 injustice

0.72

 violations

0.70

 Amendments

0.69

Activations Density 0.021%

mentions of phrases related to legal issues, particularly highlighting instances of something being deemed unconstitutional

references to unconstitutional actions or practices

No Comments

No Known Activations