INDEX

Explanations

expressions related to being out of place or out of proportion

oai_token-act-pair · gpt-3.5-turbo

expressions related to being "out of place" or "out of context."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

;;;;;;;;;;;;

-0.65

reddits

-0.63

dump

-0.61

via

-0.60

yi

-0.59

ulia

-0.59

Duo

-0.59

iaz

-0.58

 Ranking

-0.58

 partake

-0.57

POSITIVE LOGITS

 bounds

0.90

 nowhere

0.71

sted

0.71

pert

0.68

arted

0.68

formed

0.67

kil

0.67

informed

0.66

 deaf

0.65

cedented

0.64

Activations Density 0.128%

expressions related to being out of place or out of proportion

expressions related to being "out of place" or "out of context."

No Comments

No Known Activations