INDEX

Explanations

verbs related to stopping or pausing actions

oai_token-act-pair · gpt-3.5-turbo

action words related to stopping or halting processes

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

aceae

-0.90

swick

-0.73

Craft

-0.64

iverse

-0.64

 molded

-0.62

orld

-0.61

 IMAGES

-0.61

 coached

-0.60

Medium

-0.60

story

-0.59

POSITIVE LOGITS

 unilaterally

0.73

 unilateral

0.73

 moratorium

0.72

 banning

0.71

 bilateral

0.71

 bans

0.68

gag

0.68

 confisc

0.68

 compulsory

0.67

 withholding

0.67

Activations Density 0.169%

verbs related to stopping or pausing actions

action words related to stopping or halting processes

No Comments

No Known Activations