INDEX

Explanations

mentions of the word "ado" in a text

oai_token-act-pair · gpt-3.5-turbo

the suffix "ado," indicating the completion of actions or statuses

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

ancies

-0.84

ĸļ

-0.80

Ö¼

-0.80

krit

-0.80

lessly

-0.77

ologically

-0.76

utive

-0.76

orship

-0.74

itating

-0.74

ateurs

-0.73

POSITIVE LOGITS

quez

0.96

osta

0.89

 Squadron

0.86

qua

0.84

vernment

0.84

 Libre

0.83

zzi

0.80

jo

0.79

quer

0.76

quit

0.75

Activations Density 0.012%

mentions of the word "ado" in a text

the suffix "ado," indicating the completion of actions or statuses

No Comments

No Known Activations