INDEX

Explanations

sources or attributions in a document

oai_token-act-pair · gpt-3.5-turbo

references to various sources or citations in a text

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

estern

-0.83

oÄŁ

-0.76

okers

-0.71

oso

-0.69

 destro

-0.68

apo

-0.67

hma

-0.67

psey

-0.67

eg

-0.67

 cumbers

-0.65

POSITIVE LOGITS

 Sources

1.00

Fed

0.94

 Source

0.92

source

0.83

Forge

0.81

ource

0.81

Republic

0.77

books

0.76

Cub

0.74

 Republic

0.74

Activations Density 0.016%

sources or attributions in a document

references to various sources or citations in a text

No Comments

No Known Activations