INDEX

Explanations

number patterns and mathematical expressions

oai_token-act-pair · gpt-3.5-turbo

numerical values and statistics related to voting outcomes and margins

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gart

-0.66

hai

-0.65

grave

-0.65

cx

-0.63

acons

-0.62

BALL

-0.61

atha

-0.61

sha

-0.61

SX

-0.61

Reading

-0.61

POSITIVE LOGITS

 ratio

1.13

 odds

0.94

 Ratio

0.91

oldemort

0.91

 ratios

0.80

 margin

0.77

 majorities

0.73

 margins

0.72

Â½

0.72

 split

0.72

Activations Density 0.057%

number patterns and mathematical expressions

numerical values and statistics related to voting outcomes and margins

No Comments

No Known Activations