INDEX

Explanations

concepts related to conversion or transformation

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "convert" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Administration

-0.71

erk

-0.69

 Cheong

-0.67

 Epidem

-0.67

ordon

-0.65

oral

-0.63

 Anthem

-0.63

Pool

-0.63

egal

-0.62

 Eternity

-0.62

POSITIVE LOGITS

 converts

1.03

 conversions

1.00

ibles

0.99

 convert

0.98

 conversion

0.96

verting

0.96

 converted

0.94

rette

0.93

 converting

0.91

verted

0.85

Activations Density 0.009%

concepts related to conversion or transformation

occurrences of the word "convert" and its variations

No Comments

No Known Activations