INDEX

Explanations

mentions of a person named "Her"

oai_token-act-pair · gpt-3.5-turbo

references to a specific individual or character named "Her"

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

————

-0.77

ype

-0.76

ozy

-0.74

ypes

-0.74

inctions

-0.69

eering

-0.69

ornia

-0.69

anamo

-0.69

yip

-0.67

VERTIS

-0.65

POSITIVE LOGITS

itage

1.44

 Majesty

1.36

metic

1.15

ding

1.10

itability

1.05

acl

1.05

cule

1.04

self

0.97

mits

0.97

mit

0.96

Activations Density 0.064%

mentions of a person named "Her"

references to a specific individual or character named "Her"

No Comments

No Known Activations