INDEX

Explanations

names of people or entities, particularly surnames

oai_token-act-pair · gpt-3.5-turbo

words that are related to names and personal identifiers

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

 ambassadors

-0.64

 comprom

-0.62

 horizont

-0.59

 governors

-0.56

 locals

-0.53

 alleg

-0.53

 patronage

-0.53

Materials

-0.53

 independents

-0.53

 apex

-0.52

POSITIVE LOGITS

Jr

0.93

ovich

0.85

chuk

0.79

akis

0.79

(@

0.77

auer

0.77

Sr

0.76

ich

0.75

opoulos

0.75

itz

0.75

Activations Density 0.415%

names of people or entities, particularly surnames

words that are related to names and personal identifiers

No Comments

No Known Activations