INDEX

Explanations

repetitive phrases used for comparison or context in a text

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "same."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

;;;;

-0.72

*=-

-0.70

ËĪ

-0.69

ãĤ´ãĥ³

-0.68

urally

-0.68

export

-0.67

icum

-0.66

Khe

-0.65

rend

-0.64

their

-0.64

POSITIVE LOGITS

 thing

0.90

 vein

0.88

 applies

0.86

 caveats

0.84

 principle

0.84

 kind

0.80

 principles

0.79

 reasoning

0.78

 exact

0.77

 fate

0.77

Activations Density 0.042%

repetitive phrases used for comparison or context in a text

occurrences of the word "same."

No Comments

No Known Activations