INDEX

Explanations

mathematical or logical sequences

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "sequence" and its variations, indicating a focus on structured series of events or elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ryn

-0.84

aghd

-0.82

sburgh

-0.81

nai

-0.75

undai

-0.73

young

-0.72

sie

-0.71

lying

-0.70

rogens

-0.70

riel

-0.70

POSITIVE LOGITS

 Sequence

1.11

 sequence

0.94

 sequences

0.87

uality

0.79

ĸļ

0.78

 repeats

0.77

 unfold

0.75

 unfolds

0.68

 weaving

0.66

alysis

0.65

Activations Density 0.018%

mathematical or logical sequences

occurrences of the word "sequence" and its variations, indicating a focus on structured series of events or elements

No Comments

No Known Activations