INDEX

Explanations

names related to "Jal" and "Raf" in a textual context

oai_token-act-pair · gpt-3.5-turbo

names of individuals involved in legal or criminal contexts

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

ãĥĩãĤ£

-0.72

 Millennium

-0.70

 Columb

-0.69

 Dying

-0.67

 Sleeping

-0.65

Â¯Â¯

-0.64

 Prometheus

-0.64

 Founding

-0.64

%%

-0.63

ELE

-0.63

POSITIVE LOGITS

nir

1.00

Rez

0.85

icz

0.84

nic

0.84

merga

0.84

ovan

0.81

awar

0.80

utor

0.79

atility

0.79

ugal

0.78

Activations Density 0.020%

names related to "Jal" and "Raf" in a textual context

names of individuals involved in legal or criminal contexts

No Comments

No Known Activations