INDEX

Explanations

phone numbers mentioned in text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "calling."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

inth

-0.79

=-=-

-0.67

edom

-0.64

 Fiction

-0.63

imeo

-0.62

Lago

-0.61

 concentration

-0.60

rg

-0.60

eka

-0.59

 Recovery

-0.57

POSITIVE LOGITS

call

0.94

 Calling

0.92

backs

0.87

igraph

0.83

 Call

0.80

Call

0.80

calling

0.79

ipers

0.77

 calls

0.75

 forth

0.75

Activations Density 0.018%

phone numbers mentioned in text

instances of the word "calling."

No Comments

No Known Activations