INDEX

Explanations

sentences containing the word "contrary to" followed by various statements

oai_token-act-pair · gpt-3.5-turbo

phrases that contrast popular beliefs or expectations

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

estones

-0.94

onna

-0.77

esses

-0.75

oided

-0.74

ross

-0.73

olini

-0.73

leground

-0.70

hetto

-0.69

azz

-0.69

ITED

-0.67

POSITIVE LOGITS

 conventional

0.73

 prevailing

0.72

 orthodox

0.71

 altogether

0.69

ptions

0.69

ãĥĪ

0.64

 usual

0.64

 Wink

0.63

intuitive

0.63

 extant

0.62

Activations Density 0.081%

sentences containing the word "contrary to" followed by various statements

phrases that contrast popular beliefs or expectations

No Comments

No Known Activations