INDEX

Explanations

words related to representatives or representatives themselves

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "representative" and its variants

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

frey

-0.88

awar

-0.85

osure

-0.83

bows

-0.82

\\\\\\\\\\\\\\\\

-0.81

imb

-0.80

forth

-0.78

yard

-0.77

[|

-0.74

stead

-0.73

POSITIVE LOGITS

 Kislyak

0.90

hips

0.88

 warr

0.74

soType

0.73

 representatives

0.72

clinton

0.71

rored

0.70

onse

0.69

ority

0.68

holder

0.68

Activations Density 0.018%

words related to representatives or representatives themselves

occurrences of the word "representative" and its variants

No Comments

No Known Activations