INDEX

Explanations

verbs or phrases related to making changes or adjustments

oai_token-act-pair · gpt-3.5-turbo

elements associated with changes in structure or rules

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

ymph

-0.64

GV

-0.63

inar

-0.61

alone

-0.60

interstitial

-0.60

 laughs

-0.59

âĹ¼

-0.59

heny

-0.58

Bei

-0.57

contained

-0.57

POSITIVE LOGITS

 accordingly

1.24

 drastically

1.22

 dramatically

1.09

 radically

1.08

 unilaterally

0.88

 substantially

0.86

 forever

0.82

 overnight

0.81

 subtly

0.80

 abruptly

0.80

Activations Density 0.241%

verbs or phrases related to making changes or adjustments

elements associated with changes in structure or rules

No Comments

No Known Activations