INDEX

Explanations

words ending in 'iter'

oai_token-act-pair · gpt-3.5-turbo

instances of the word "iter" or related terms that suggest ongoing actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ership

-0.79

¥µ

-0.73

ruciating

-0.72

¶æ

-0.66

 bottled

-0.65

itionally

-0.65

ecause

-0.62

sets

-0.61

¬¼

-0.61

Bay

-0.60

POSITIVE LOGITS

iter

0.94

ariat

0.84

aton

0.84

mination

0.78

ators

0.78

itone

0.75

arity

0.75

ati

0.74

asury

0.73

imation

0.72

Activations Density 0.010%

words ending in 'iter'

instances of the word "iter" or related terms that suggest ongoing actions or processes

No Comments

No Known Activations