INDEX

Explanations

names mentioned in various contexts

oai_token-act-pair · gpt-3.5-turbo

the mention of a particular name or entity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

rador

-0.85

inarily

-0.78

hips

-0.74

 bearer

-0.74

rican

-0.71

rament

-0.71

displayText

-0.71

IAL

-0.68

 asses

-0.68

 glim

-0.68

POSITIVE LOGITS

arest

1.13

braska

1.00

gan

0.91

lde

0.89

cht

0.89

jad

0.87

cker

0.84

zel

0.83

verend

0.83

ema

0.83

Activations Density 0.016%

names mentioned in various contexts

the mention of a particular name or entity

No Comments

No Known Activations