INDEX

Explanations

phrases related to actions and instructions

oai_token-act-pair · gpt-3.5-turbo

actions and activities in the narrative context

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

unity

-0.74

cens

-0.73

ection

-0.72

chev

-0.70

ortium

-0.70

uum

-0.69

avin

-0.66

][

-0.66

Ã£

-0.64

Ã¡s

-0.63

POSITIVE LOGITS

 alike

0.80

 frantically

0.66

instead

0.65

 blindly

0.65

 prest

0.65

 anew

0.63

 cryst

0.61

Doc

0.61

 REPL

0.61

 immediately

0.60

Activations Density 0.376%

phrases related to actions and instructions

actions and activities in the narrative context

No Comments

No Known Activations