INDEX

Explanations

words or phrases related to evaluating the impact of actions or situations

oai_token-act-pair · gpt-3.5-turbo

phrases expressing benefits or positives for various subjects or contexts

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

aque

-0.77

rette

-0.68

ohn

-0.67

aram

-0.67

venants

-0.67

cles

-0.67

 laun

-0.66

rose

-0.65

clone

-0.65

operated

-0.65

POSITIVE LOGITS

 morale

0.93

 everybody

0.86

geries

0.83

gotten

0.78

 laughs

0.78

us

0.77

 storing

0.76

 awhile

0.76

 everyone

0.76

 efficiency

0.75

Activations Density 0.097%

words or phrases related to evaluating the impact of actions or situations

phrases expressing benefits or positives for various subjects or contexts

No Comments

No Known Activations