INDEX

Explanations

names of specific companies or products

oai_token-act-pair · gpt-3.5-turbo

mentions of "Land" and related contexts

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

 whistle

-0.83

 postal

-0.78

 candle

-0.76

 Cantor

-0.75

 stripes

-0.73

box

-0.72

 razor

-0.72

 chop

-0.71

Pis

-0.70

 thorough

-0.69

POSITIVE LOGITS

Land

2.71

Med

1.58

Fuel

1.54

Park

1.40

LAND

1.36

Monitor

1.36

Research

1.26

Mech

1.25

Deploy

1.25

fuel

1.25

Activations Density 0.044%

names of specific companies or products

mentions of "Land" and related contexts

No Comments

No Known Activations