INDEX

Explanations

instances of uncertainty or possibility

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "may."

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

ging

-0.72

feeding

-0.65

 Fighter

-0.65

hered

-0.64

rency

-0.62

ciating

-0.62

ãĥ¼ãĥ«

-0.61

raphic

-0.61

performing

-0.61

lier

-0.60

POSITIVE LOGITS

hap

1.10

haps

1.09

onna

1.07

be

0.94

bes

0.88

 derive

0.87

 confuse

0.85

 differ

0.83

 optionally

0.83

 misunder

0.82

Activations Density 0.071%

instances of uncertainty or possibility

occurrences of the word "may."

No Comments

No Known Activations