INDEX

Explanations

statistical information or numerical references

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate statistical estimates or proportions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

umbn

-0.75

ffen

-0.61

CV

-0.61

ilic

-0.57

agate

-0.56

BAT

-0.55

ãĤ¹ãĥĪ

-0.55

ohan

-0.55

ensis

-0.55

cart

-0.55

POSITIVE LOGITS

 half

1.26

1.15

 eighty

1.09

1.07

1.06

 ninety

1.06

 seventy

1.05

1.04

1.03

Activations Density 0.094%

statistical information or numerical references

phrases that indicate statistical estimates or proportions

No Comments

No Known Activations