INDEX

Explanations

studies or reports revealing statistical findings or survey results

oai_token-act-pair · gpt-3.5-turbo

phrases indicating research findings or survey results

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

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.75

ãĤ§

-0.65

via

-0.64

ï¸

-0.63

ä½ľ

-0.62

vard

-0.61

ãĤ¶

-0.61

argon

-0.60

ãĥ¤

-0.60

gur

-0.60

POSITIVE LOGITS

Orb

0.64

 discrepancies

0.63

 anomalies

0.62

evidence

0.61

¿½

0.61

 irregularities

0.59

 citing

0.57

 findings

0.56

agi

0.56

 that

0.56

Activations Density 0.220%

studies or reports revealing statistical findings or survey results

phrases indicating research findings or survey results

No Comments

No Known Activations