INDEX

Explanations

mentions of specific individuals and their actions

oai_token-act-pair · gpt-3.5-turbo

possessive forms and references to individuals or entities

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

Reviewer

-0.80

hari

-0.80

KT

-0.79

livion

-0.76

Quantity

-0.75

ENN

-0.75

BT

-0.75

Ú

-0.74

xxx

-0.74

>]

-0.73

POSITIVE LOGITS

 newest

1.06

 biggest

1.06

own

0.96

 inability

0.95

 youngest

0.94

 oldest

0.92

 latest

0.92

 attorney

0.91

 penchant

0.90

 greatest

0.89

Activations Density 0.130%

mentions of specific individuals and their actions

possessive forms and references to individuals or entities

No Comments

No Known Activations