INDEX

Explanations

the word "intended," especially when associated with specific actions or outcomes

oai_token-act-pair · gpt-3.5-turbo

expressions of intention or purpose

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Solitaire

-0.76

acks

-0.63

anca

-0.62

adier

-0.58

Sort

-0.57

Du

-0.57

aura

-0.57

Americ

-0.57

 ranked

-0.57

Plus

-0.57

POSITIVE LOGITS

lessly

0.81

 Parenthood

0.79

fully

0.75

ãĥĨ

0.69

llor

0.68

 provoking

0.68

only

0.66

ller

0.66

nance

0.65

ful

0.65

Activations Density 0.044%

the word "intended," especially when associated with specific actions or outcomes

expressions of intention or purpose

No Comments

No Known Activations