INDEX

Explanations

words related to wheels and wheel-related activities

oai_token-act-pair · gpt-3.5-turbo

references to wheels and wheel-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Suc

-0.77

 Grounds

-0.72

 Diaz

-0.69

 Wong

-0.68

uates

-0.67

 Sark

-0.66

 Gleaming

-0.66

ABE

-0.65

 Norn

-0.64

 Tonight

-0.64

POSITIVE LOGITS

chairs

1.48

chair

1.40

wright

1.21

wheel

1.10

bar

1.09

base

1.03

house

0.99

loader

0.97

 bearings

0.92

wagon

0.90

Activations Density 0.045%

words related to wheels and wheel-related activities

references to wheels and wheel-related concepts

No Comments

No Known Activations