INDEX

Explanations

mentions or descriptions of tents

oai_token-act-pair · gpt-3.5-turbo

references to tents or similar structures

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

hire

-0.69

onso

-0.69

history

-0.64

iano

-0.64

ills

-0.63

javascript

-0.63

 Wayne

-0.63

andro

-0.62

users

-0.62

 Schro

-0.62

POSITIVE LOGITS

 tent

4.01

 tents

2.70

 Tent

2.59

 encamp

1.57

 tentacles

1.46

 camp

1.41

 camps

1.30

hut

1.26

 camping

1.17

RV

1.04

Activations Density 0.016%

mentions or descriptions of tents

references to tents or similar structures

No Comments

No Known Activations