INDEX

Explanations

variations of the word "jam."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_6.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.6.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Recall

-0.67

 Highlander

-0.61

 erection

-0.60

 Aires

-0.59

 escal

-0.59

avement

-0.58

 Equality

-0.58

ģ«

-0.57

 Marketable

-0.57

gio

-0.57

POSITIVE LOGITS

monds

1.11

mond

1.01

bell

1.00

oon

0.99

mer

0.98

ming

0.97

pering

0.96

pered

0.92

pled

0.92

puter

0.91

Activations Density 0.030%

variations of the word "jam."

No Comments

No Known Activations