INDEX

Explanations

specific mention of tables

oai_token-act-pair · gpt-3.5-turbo

references to tables in documents

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

vernment

-0.74

qua

-0.71

 holiest

-0.66

=-=-=-=-=-=-=-=-

-0.65

 Monetary

-0.65

 Advocate

-0.63

Â¯

-0.62

 Patt

-0.61

 Chosen

-0.61

 protective

-0.60

POSITIVE LOGITS

cloth

1.19

sheet

0.95

poons

0.90

au

0.89

table

0.88

Table

0.88

tops

0.87

sheets

0.81

top

0.81

poon

0.78

Activations Density 0.014%

specific mention of tables

references to tables in documents

No Comments

No Known Activations