INDEX

Explanations

phrases related to investigations

oai_token-act-pair · gpt-3.5-turbo

mentions of investigations

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

emonic

-0.80

licks

-0.72

birth

-0.70

yip

-0.64

 maiden

-0.63

generic

-0.61

ixed

-0.61

absolute

-0.61

aez

-0.59

-+

-0.59

POSITIVE LOGITS

 investigation

0.90

 probing

0.86

 Investigation

0.84

 investigates

0.82

 Probe

0.81

 involving

0.81

 investigating

0.80

 inquiry

0.80

 investigations

0.79

 undertaken

0.75

Activations Density 0.026%

phrases related to investigations

mentions of investigations

No Comments

No Known Activations