INDEX

Explanations

phrases or sentences with intensifiers and actions escalating towards a critical point

oai_token-act-pair · gpt-3.5-turbo

phrases indicating extreme situations or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

akia

-0.58

rounder

-0.58

odge

-0.57

 DRAG

-0.57

 Together

-0.56

 Others

-0.56

missions

-0.55

asts

-0.55

 Variant

-0.54

olly

-0.52

POSITIVE LOGITS

 where

1.07

 whereby

0.83

 whence

0.80

 point

0.79

lessness

0.78

 that

0.76

manship

0.74

ophys

0.74

 brink

0.72

 wherein

0.71

Activations Density 0.041%

phrases or sentences with intensifiers and actions escalating towards a critical point

phrases indicating extreme situations or conditions

No Comments

No Known Activations